From b9a3c7df1ec9497014886af2af967272b5af2d9c Mon Sep 17 00:00:00 2001 From: Joy Arulraj Date: Thu, 5 Oct 2023 11:25:53 -0400 Subject: [PATCH] updates --- docs/source/reference/optimizations.rst | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/source/reference/optimizations.rst b/docs/source/reference/optimizations.rst index 5e9b2c5a8f..9a67e1da1d 100644 --- a/docs/source/reference/optimizations.rst +++ b/docs/source/reference/optimizations.rst @@ -5,10 +5,10 @@ EvaDB Optimizations 🛠️ EvaDB optimizes the evaluation of AI functions using these optimizations: -1️⃣ Result Caching: EvaDB caches outcomes from expensive function invocations during query processing. This approach facilitates faster retrieval in subsequent queries. 📂 +1️⃣ Result Caching: EvaDB caches outcomes from expensive function invocations during query processing. This approach facilitates faster retrieval in subsequent queries. 📂⚡ -2️⃣ Predicate Reordering: Efficiency is key. EvaDB strategically reorders predicates to prioritize lower-cost and more selective evaluations. ⚖️ +2️⃣ Predicate Reordering: Efficiency is key. EvaDB strategically reorders predicates to prioritize lower-cost and more selective evaluations. 🔀🕰️ -3️⃣ Parallel Processing with Ray: Leveraging the Ray framework, EvaDB runs AI models in parallel, optimizing GPU utilization. Additionally, an AI pipeline is established for concurrent CPU tasks, such as data loading and decoding. 🚀 +3️⃣ Parallel Processing with Ray: Leveraging the Ray framework, EvaDB runs AI models in parallel, optimizing GPU utilization. Additionally, an AI pipeline is established for concurrent CPU tasks, such as data loading and decoding. 🚄🎩 -These techniques ensure superior performance and responsiveness in EvaDB's AI function evaluations. +These techniques ensure superior performance and responsiveness in EvaDB's AI function evaluations. Dive in and experience the EvaDB difference! 🌟🎉